molecules

Article

Chemical Synthesis ofDeoxynivalenol-3-β-D-[13C6]-glucoside andApplication in Stable Isotope Dilution Assays

Katharina Habler 1, Oliver Frank 2 and Michael Rychlik 1,*1 Analytical Food Chemistry, Technische Universität München, Alte Akademie 10, D-85354 Freising, Germany;

katharina.habler@tum.de2 Food Chemistry and Molecular Sensory Science, Technische Universität München, Lise-Meitner-Straße 34,

D-85354 Freising, Germany; oliver.frank@tum.de* Correspondence: michael.rychlik@tum.de; Tel.: +49-8161-71-3746

Academic Editor: Mary FletcherReceived: 30 May 2016; Accepted: 20 June 2016; Published: 27 June 2016

Abstract: Modified mycotoxins have been gaining importance in recent years and present acertain challenge in LC-MS/MS analysis. Due to the previous lack of a labeled isotopologueof the modified mycotoxin deoxynivalenol-3-glucoside, in our study we synthesized the first13C-labeled internal standard. Therefore, we used the Königs-Knorr method to synthesizedeoxynivalenol-3-β-D-[13C6]-glucoside originated from unlabeled deoxynivalenol and [13C6]-labeledglucose. Using the synthesized isotopically-labeled standard deoxynivalenol-3-β-D-[13C6]-glucosideand the purchased labeled standard [13C15]-deoxynivalenol, a stable isotope dilution LC-MS/MSmethod was firstly developed for deoxynivalenol-3-glucoside and deoxynivalenol in beer. Thepreparation and purification of beer samples was based on a solid phase extraction. The validationdata of the newly developed method gave satisfying results. Intra- and interday precision studiesrevealed relative standard deviations below 0.5% and 7%, respectively. The recoveries ranged forboth analytes between 97% and 112%. The stable isotope dilution assay was applied to various beersamples from four different countries. In summary, deoxynivalenol-3-glucoside and deoxynivalenolmostly appeared together in varying molar ratios but were quantified in rather low contents in theinvestigated beers.

Keywords: modified mycotoxin; deoxynivalenol-3-glucoside; stable isotope dilution assay;LC-MS/MS; beer; labeled standard; synthesis

1. Introduction

Fusarium species are ubiquitously present and can infect a wide range of crops. Fusarium headblight, a well-known crop disease caused by different Fusarium species, like F. graminearum, F. culmorum,or F. avenaceum [1], can lead to yield loss and reduce grain quality. Fungal infestation of cereals, suchas maize, wheat, and barley, is often associated with mycotoxin contamination and, hence, especiallyaffects the safety of human and animal diets. Fusarium toxins can be categorized in several majorgroups: fumonisins, enniatins, zearalenones, and trichothecenes [2]. The latter group can be divided intype A, B, C, and D trichothecenes, among which type A and B toxins are most abundant and relevantin food and feed [2]. Deoxynivalenol (DON) and its modified form deoxynivalenol-3-glucoside (D3G)(Figure 1) rank among the best-known representatives of type B trichothecenes and can present aserious health risk for humans and animals [2].

Molecules 2016, 21, 838; doi:10.3390/molecules21070838 www.mdpi.com/journal/molecules

Molecules 2016, 21, 838 2 of 15

Molecules 2016, 21, 838 2 of 15

derivatives was established, but D3G could not be included because of insufficient toxicity and exposure data [4].

More than a decade ago D3G was identified as the most important metabolite of DON and, later, could be found in naturally-contaminated wheat and maize [5,6]. Due to phase-II-metabolism in plants DON can be metabolized and end up as D3G. Moreover, hydrolysis of D3G in human and animal gastrointestinal tracts can again release the aglycone DON [7–9], which additionally contributes to the base contamination and plays an emerging issue concerning food safety and health risk assessments. In addition to D3G, several other modified metabolites have already been reported, for example zearalenone-glucoside and -sulfate, nivalenol-glucoside, deoxynivalenol-sulfate, deoxynivalenol-cysteine conjugates, deoxynivalenol-oligo/poly-glucosides, or deoxynivalenol-glutathione adducts [10–17].

Modified mycotoxins are beyond the normal scope of routine analysis, thus leaving them often disregarded unintentionally. However, D3G has already been observed in cereals and barley malt in concentrations up to 19,000 μg/kg [18–22]. In beer samples, the levels of D3G often exceed the levels of DON [23,24] and underlines the urgent need for a substantial risk assessment. Due to the lack of a stable isotope-labeled internal standard, previous studies often quantified D3G by matrix-matched calibration [18–20,24]. However, it is generally accepted that isotopologue standards are the tools of choice to compensate for matrix interferences during LC-MS/MS measurements [25].

Due to the lack of a stable isotopologue, the aim of our study was the chemical synthesis of an isotopically-labeled standard of D3G by means of the Königs-Knorr method. We developed a stable isotope dilution assay (SIDA) for D3G and DON by using the newly-synthesized internal standard deoxynivalenol-3-β-D-[13C6]-glucoside and the commercially available standard [13C15]-DON, respectively. The thoroughly validated LC-MS/MS method was applied to various beer samples.

Figure 1. Structures of D3G and DON.

2. Results

2.1. Syntheses

2.1.1. Synthesis of 2,3,4,6-Tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside

The acetylation of [13C6]-glucose was performed with acetic anhydride and pyridine according to Asam and Rychlik [26]. A complete acetylation resulted after 48 h, giving a lightly yellow and clear solution. After a washing step with brine and removal of the solvent, the product 1,2,3,4,6-pentaacetyl-D-[13C6]-glucopyranoside crystallized during evaporation under cooling in a yield of 87% (0.96 g, 2.4 mmol).

Bromination of the acetylated [13C6]-glucose (0.5 g, 1.3 mmol) was accomplished with hydrobromic acid in acetic acid following a procedure reported by Ravindranathan Kartha and Jennings, Redemann and Niemann, and Koschella et al. [27–29]. The crude product was extracted with dichloromethane followed by successively washing with aqueous sodium thiosulfate, aqueous sodium hydrogen carbonate, and brine. After dissolving the product in diethyl ether and the addition of hexane 2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside crystallized under cooling as a white compound, yielding 95% (0.5 g, 1.2 mmol).

Figure 1. Structures of D3G and DON.

Maximum limits for unprocessed cereals are legislatively set for DON and zearalenone at1250 µg/kg and 100 µg/kg, respectively, but have not yet been defined for modified mycotoxins, likeD3G [3]. A provisional tolerable daily intake of 1 µg/kg body weight for the sum of DON and itsacetylated derivatives was established, but D3G could not be included because of insufficient toxicityand exposure data [4].

More than a decade ago D3G was identified as the most important metabolite of DON and,later, could be found in naturally-contaminated wheat and maize [5,6]. Due to phase-II-metabolismin plants DON can be metabolized and end up as D3G. Moreover, hydrolysis of D3G inhuman and animal gastrointestinal tracts can again release the aglycone DON [7–9], whichadditionally contributes to the base contamination and plays an emerging issue concerning foodsafety and health risk assessments. In addition to D3G, several other modified metaboliteshave already been reported, for example zearalenone-glucoside and -sulfate, nivalenol-glucoside,deoxynivalenol-sulfate, deoxynivalenol-cysteine conjugates, deoxynivalenol-oligo/poly-glucosides, ordeoxynivalenol-glutathione adducts [10–17].

Modified mycotoxins are beyond the normal scope of routine analysis, thus leaving them oftendisregarded unintentionally. However, D3G has already been observed in cereals and barley malt inconcentrations up to 19,000 µg/kg [18–22]. In beer samples, the levels of D3G often exceed the levelsof DON [23,24] and underlines the urgent need for a substantial risk assessment. Due to the lack of astable isotope-labeled internal standard, previous studies often quantified D3G by matrix-matchedcalibration [18–20,24]. However, it is generally accepted that isotopologue standards are the tools ofchoice to compensate for matrix interferences during LC-MS/MS measurements [25].

Due to the lack of a stable isotopologue, the aim of our study was the chemical synthesis ofan isotopically-labeled standard of D3G by means of the Königs-Knorr method. We developeda stable isotope dilution assay (SIDA) for D3G and DON by using the newly-synthesizedinternal standard deoxynivalenol-3-β-D-[13C6]-glucoside and the commercially available standard[13C15]-DON, respectively. The thoroughly validated LC-MS/MS method was applied to variousbeer samples.

2. Results

2.1. Syntheses

2.1.1. Synthesis of 2,3,4,6-Tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside

The acetylation of [13C6]-glucose was performed with acetic anhydride and pyridine accordingto Asam and Rychlik [26]. A complete acetylation resulted after 48 h, giving a lightly yellowand clear solution. After a washing step with brine and removal of the solvent, the product1,2,3,4,6-pentaacetyl-D-[13C6]-glucopyranoside crystallized during evaporation under cooling in ayield of 87% (0.96 g, 2.4 mmol).

Molecules 2016, 21, 838 3 of 15

Bromination of the acetylated [13C6]-glucose (0.5 g, 1.3 mmol) was accomplished withhydrobromic acid in acetic acid following a procedure reported by Ravindranathan Kartha andJennings, Redemann and Niemann, and Koschella et al. [27–29]. The crude product was extracted withdichloromethane followed by successively washing with aqueous sodium thiosulfate, aqueous sodiumhydrogen carbonate, and brine. After dissolving the product in diethyl ether and the addition ofhexane 2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside crystallized under cooling as a whitecompound, yielding 95% (0.5 g, 1.2 mmol).

To obtain enough substance of 2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside (in total1.6 g) for the following synthesis of deoxynivalenol-3-β-D-[13C6]-glucoside, the acetylation andbromination syntheses were repeated four times, respectively.

2.1.2. Synthesis of Deoxynivalenol-3-β-D-[13C6]-glucoside

Analogously to Savard et al. [30] and Mikula et al. [31,32], for the synthesis ofdeoxynivalenol-3-β-D-[13C6]-glucoside the Königs-Knorr method was used. Therefore, theeducts DON and 2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside were mixed with thecatalyst silver carbonate in dichloromethane. To facilitate the reaction a surplus of2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside and silver carbonate were necessary and hence,were added two further times. Surplus silver ions were precipitated with potassium thiocyanatesolution and severed by membrane filtration after reaction.

The controlled hydrolysis of deoxynivalenol-3-2,3,4,6-tetra-O-acetyl-β-D-[13C6]-glucoside todeoxynivalenol-3-β-D-[13C6]-glucoside was performed in 5 mL acetonitrile/water (1/1, v/v) withpotassium hydroxide. The stoichiometrically-required amount of potassium hydroxide was addeddropwise under continuous pH control via pH meter. Due to instability of type B trichothecenesin alkaline range [33], the addition of potassium hydroxide took place drop by drop all day longto guarantee a pH value below eleven. On the next day the hydrolysis reaction was stopped byneutralization with hydrochloride acid before preparative purification.

The synthesis procedure was performed 72 times to obtain a sufficient substance amount asupscaling studies of the reaction resulted in a lower yield.

The recorded 1H-NMR and 13C-NMR spectra of deoxynivalenol-3-β-D-[13C6]-glucoside show thefollowing signals:

1H-NMR (500 MHz, deuterium oxide) δ 6.65 (dd, J = 5.8, 1.3 Hz, 1H, H-10), 4.95 (d, J = 5.8 Hz, 1H,H-11), 4.54 (dd, J = 10.9, 4.4 Hz, 1H, H-3), 5.00–4.00 (s, 1H, H-7, d, 1H, H-2), 5.00–3.00 (m, 7H, H-11,H-21, H-31, H-41, H-51, H-61a, H-61b), 4.00–3.00 (d, 1H, H-13a, d, 1H, H-13b, d, 1H, H-15a, d, 1H, H-15b),2.50–1.50 (dd, 1H, H-4a, dd, 1H, H-4b), 1.91 (s, 3H, -CH3-16), 1.12 (s, 3H, -CH3-14).

13C-NMR (125 MHz, deuterium oxide) δ 202.2 (1C, C-8), 138.1 (1C, C-9), 137.0 (1C, C-10), 97.4 (1C,C-11, m), 82.0–60.4 (6C, C-3, C-2, C-4, C-7, C-11, C-15), 75.1–76.4 (2C, C-31, C-51, m) 73.0 (1C, C-21, m),69.5(1C, C-41, m), 60.7 (1C, C-61, m), 46.0 (1C, C-13), 14.6 (1C, C-16), 13.9 (1C, C-14).

NMR experiments of deoxynivalenol-3-β-D-[13C6]-glucoside were performed only with a focuson assigning the glycosidic bond between C-3 of DON and C-11 of glucose, which could be verified.For quantitation via qNMR [34] the signal of deoxynivalenol-3-β-D-[13C6]-glucoside at 6.65 ppm (H-10)was chosen. The molar concentration was 0.6106 mmol/L. The yield was 1% (170 µg, 0.37 µmol).

LC-MS/MS fragmentation spectra of D3G and [13C6]-D3G are displayed in Figure 2. LC-MS/MSmeasurements were carried out in the negative ESI mode and the respective [M ´ H]´ ions asprecursor ions were used. D3G (m/z 457.3) and [13C6]-D3G (m/z 463.3) likewise showed the loss of the-CH2OH group at position C-6 of the DON molecule resulting in the most intensive fragment for D3G(m/z 427.2 = 457.3 ´ 30) and [13C6]-D3G (m/z 433.2 = 463.3 ´ 30) as Berthiller et al. [35] alreadyreported for D3G. A further loss of water and glucose or [13C6]-glucose followed, respectively(D3G m/z 247.3 = 457.3 ´ 30 ´ 18 ´ 162, [13C6]-D3G m/z 247.3 = 463.3 ´ 30 ´ 18 ´ 168). For D3G, thetransitions (m/z 457.3Ñ 427.2) and (m/z 457.3Ñ 247.3), and for [13C6]-D3G the transitions (m/z 463.3Ñ 433.2) and (m/z 463.3Ñ 247.3), served as quantifier and qualifier, respectively.

Molecules 2016, 21, 838 4 of 15

Molecules 2016, 21, 838 4 of 15

of D3G and DON could be assured by using a Hydrosphere RP-C18 column (YMC Europe GmbH, Dinslaken, Germany) combined with a specific gradient.

Figure 2. LC-MS/MS fragmentation spectra of D3G and [13C6]-D3G.

Table 1. LC-MS/MS parameter of the stable isotope dilution assay of D3G and DON.

Analyte Precursor Ion m/z Product Ion m/z Q1 Pre Bias (V) Q3 Pre Bias (V) CE (V)

D3G 457.3 427.2 17 13 16 247.3 22 15 22

[13C6]-D3G 463.3 433.2 17 13 16 247.3 22 15 22

DON 297.2 249.2 −21 −18 −12 231.2 −20 −17 −12

[13C15]-DON 312.2 265.2 −21 −18 −12 247.2 −20 −17 −12

2.2.2. Calibration and Quantitation

Response functions were obtained using linear regression. The response factors were 1.1 for D3G and DON, respectively. The coefficients of determination (R2) exceeded 0.9994 and as confirmed by Mandel test, the calibration curves showed linearity for D3G and DON within the molar ratios 0.1–10, respectively.

2.2.3. Sample Purification

The purification of beer samples was based on a solid phase extraction. Due to the already developed multi-mycotoxin method of cereals [18], the application of Bond Elut Mycotoxin cartridges (Agilent Technologies, Santa Clara, CA, USA) were again selected.

A ratio of 0.8 mL beer and 4 mL acetonitrile was chosen to precipitate polar matrix compounds in beer and to follow the recommendations of the manufacturer for loading the cartridge [36,37]. Polar matrix compounds in beer were precipitated after mixing the chosen beer and an acetonitrile volume of 0.8 mL and 4 mL, respectively. An overload of the cartridges was not observed with a total sample volume of 4.8 mL. After evaporation of the eluates until dryness and reconstitution of the sample with 200 μL acetonitrile/water (1/1, v/v) a lightly yellow colored clear sample resulted. The LC-MS/MS chromatograms of beer samples revealed low noise signals that did not interfere or overlap with the signals of the analytes and labeled standards. Figure 3 shows a LC-MS/MS

Figure 2. LC-MS/MS fragmentation spectra of D3G and [13C6]-D3G.

2.2. Method Development

2.2.1. LC-MS/MS

DON and [13C15]-DON were measured in the positive ESI mode. Therefore, the protonatedmolecules were used as precursor ions, respectively. D3G was analyzed in the negative ESI modeand the abundant [M ´ H]´ ion was used as precursor ion. As already described, the labeledstandard [13C6]-D3G gave a fragmentation pattern similar to the respective unlabeled compound.The LC-MS/MS parameter of the SIDA of D3G and DON are shown in Table 1. A chromatographicseparation of D3G and DON could be assured by using a Hydrosphere RP-C18 column (YMC EuropeGmbH, Dinslaken, Germany) combined with a specific gradient.

Table 1. LC-MS/MS parameter of the stable isotope dilution assay of D3G and DON.

Analyte Precursor Ion m/z Product Ion m/z Q1 Pre Bias (V) Q3 Pre Bias (V) CE (V)

D3G 457.3427.2 17 13 16

247.3 22 15 22

[13C6]-D3G 463.3433.2 17 13 16

247.3 22 15 22

DON 297.2249.2 ´21 ´18 ´12

231.2 ´20 ´17 ´12

[13C15]-DON 312.2265.2 ´21 ´18 ´12

247.2 ´20 ´17 ´12

2.2.2. Calibration and Quantitation

Response functions were obtained using linear regression. The response factors were 1.1 forD3G and DON, respectively. The coefficients of determination (R2) exceeded 0.9994 and as confirmedby Mandel test, the calibration curves showed linearity for D3G and DON within the molar ratios0.1–10, respectively.

2.2.3. Sample Purification

The purification of beer samples was based on a solid phase extraction. Due to the alreadydeveloped multi-mycotoxin method of cereals [18], the application of Bond Elut Mycotoxin cartridges(Agilent Technologies, Santa Clara, CA, USA) were again selected.

Molecules 2016, 21, 838 5 of 15

A ratio of 0.8 mL beer and 4 mL acetonitrile was chosen to precipitate polar matrix compounds inbeer and to follow the recommendations of the manufacturer for loading the cartridge [36,37]. Polarmatrix compounds in beer were precipitated after mixing the chosen beer and an acetonitrile volumeof 0.8 mL and 4 mL, respectively. An overload of the cartridges was not observed with a total samplevolume of 4.8 mL. After evaporation of the eluates until dryness and reconstitution of the sample with200 µL acetonitrile/water (1/1, v/v) a lightly yellow colored clear sample resulted. The LC-MS/MSchromatograms of beer samples revealed low noise signals that did not interfere or overlap with thesignals of the analytes and labeled standards. Figure 3 shows a LC-MS/MS chromatogram of D3Gand DON as well as their labeled standards [13C6]-D3G and [13C15]-DON in a naturally-contaminatedbeer sample.

Molecules 2016, 21, 838 5 of 15

chromatogram of D3G and DON as well as their labeled standards [13C6]-D3G and [13C15]-DON in a naturally-contaminated beer sample.

Figure 3. LC-MS/MS chromatograms of (a) D3G in negative ESI mode; and (b) DON in positive ESI mode in a naturally-contaminated beer sample with 10.6 μg/L D3G and 20.3 μg/L DON.

2.3. Method Validation

2.3.1. Limits of Detection and Quantitation

The LODs and LOQs for the SIDAs were calculated according to Vogelgesang and Hädrich [38]. Beer free of the monitored mycotoxins was used as blank matrix. The LOD and LOQ were 2.99 μg/L and 8.84 μg/L for D3G and 1.49 μg/L and 4.44 μg/L for DON, respectively (Table 2).

Figure 3. LC-MS/MS chromatograms of (a) D3G in negative ESI mode; and (b) DON in positive ESImode in a naturally-contaminated beer sample with 10.6 µg/L D3G and 20.3 µg/L DON.

Molecules 2016, 21, 838 6 of 15

2.3. Method Validation

2.3.1. Limits of Detection and Quantitation

The LODs and LOQs for the SIDAs were calculated according to Vogelgesang and Hädrich [38].Beer free of the monitored mycotoxins was used as blank matrix. The LOD and LOQ were 2.99 µg/Land 8.84 µg/L for D3G and 1.49 µg/L and 4.44 µg/L for DON, respectively (Table 2).

Table 2. Validation data of the stable isotope dilution assay of D3G and DON.

Analyte LOD(µg/L)

LOQ(µg/L)

Precision (n = 3) (RSD %) Recoveries (%) at Spiking Levels of

Intraday Interday 10 µg/L 20 µg/L 30 µg/L

D3G 2.99 8.84 0.5 5 112 ˘ 3 110 ˘ 3 105 ˘ 1DON 1.49 4.44 0.3 7 97 ˘ 1 107 ˘ 2 108 ˘ 2

2.3.2. Recovery

The recoveries of the SIDAs were determined at three different spiking levels for each mycotoxin.The recoveries ranged between 112% and 105% for D3G and between 97% and 108% for DON withrelative standard deviations (RSDs) below 3% (Table 2).

2.3.3. Precision

The intra-day (n = 3) and inter-day (n = 3) coefficients of variation are shown in Table 2. Theintra-day precision ranged between 0.3% and 0.5% and the inter-day precision ranged between 5%and 7%.

2.4. Analysis of D3G and DON in Beer Samples

Different beer samples bought in 2015 from a local market in Munich were analyzed with theSIDA of D3G and DON presented. The nine different beers originated from four different countries,namely Germany, Taiwan, China, and the USA. Apart from one organic wheat beer with 5.1 vol %alcohol, the other conventional beers had an alcohol content varying between 4.7 vol % and 8.3 vol %.The origins, alcohol contents, and the contents of D3G and DON of the analyzed beer samples aresummarized in Table 3. On average, the investigated beers were contaminated with 22.6 µg/L D3Gand 13.3 µg/L DON. D3G was found with concentrations between 8.31 µg/L and 63.3 µg/L in sixbeer samples. Seven of the nine analyzed beer samples were contaminated with DON between5.11 µg/L and 28.3 µg/L. The molar ratios between DON and D3G ranged between 0.34 and 1.54.

Table 3. Contents of D3G and DON in the analyzed beer samples.

Beer Type Country Alcohol Content(vol %) D3G (µg/L) DON (µg/L) Molar Ratio

n(D3G)/n(DON)

Lager Germany 5.4 - - -Wheat beer b Germany 5.1 10.6 20.3 0.34

Lager Taiwan 5.0 15.3 6.43 1.54Lager China 4.7 8.31 a 5.11 1.05Lager USA 5.0 20.1 14.1 0.92IPA USA 6.5 - 9.77 -

Triple Golden Ale USA 8.0 18.1 9.40 1.25Porter USA 5.3 - - -

Belgian Style IPA USA 8.3 63.3 28.3 1.45

- not detected; a below LOQ; b organic; IPA, Indian Pale Ale.

Molecules 2016, 21, 838 7 of 15

3. Discussion

3.1. Syntheses

In addition to the Königs-Knorr method [39], the Schmidt method [40] also offersa possible synthetic route for glycosylations. In our study, we decided to synthesizedeoxynivalenol-3-β-D-[13C6]-glucoside according to the Königs-Knorr synthesis based on the eductsDON and 2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside with the catalyst silver carbonatein dichloromethane.

The synthesis of 2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside provided an easyreaction in very good yield above 85%. In contrast, the glycosylation reaction and especiallythe hydrolysis of the acetyl groups presented a bigger challenge. Additionally, dichloromethane,acetonitrile as an organic solvent, as well as different catalysts like silver carbonate, silver oxide,and silver triflate were tested. All reaction experiments revealed that no product was formed withsilver triflate independent of the organic solvent. The synthesis was also performed with molecularsieves (3 and 4 Å) and under argon atmosphere again yielding no product. Better yields wereachieved by using silver carbonate and dichloromethane instead of silver oxide and acetonitrile.However, a crucial aspect for higher yields turned out to be the increase of the equivalents of2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside and silver carbonate from two to twenty,compared to DON.

Due to the instability of DON under alkaline conditions [33], several hydrolysis variationswere considered [30–32,41]: potassium carbonate in methanol; potassium cyanide in methanol;potassium hydroxide in water, methanol, tetrahydrofuran/water, and acetonitrile/water. Finally, thedeprotection of the acetylated product deoxynivalenol-3-2,3,4,6-tetra-O-acetyl-β-D-[13C6]-glucosideto deoxynivalenol-3-β-D-[13C6]-glucoside under continuous pH-value control via pH meter withpotassium hydroxide in acetonitrile/water emerged as the most practicable solution.

An option to prevent the formation of deoxynivalenol-15-β-D-[13C6]-glucoside,15-acetyldeoxynivalenol with its hydroxyl group at C-15 being protected could have beenused as the educt. However, as the hydrolysis of the acetyl groups has already presented a problem, itappeared more reasonable to purify the synthesis and separate [13C6]-D3G from [13C6]-D15G, DON,and other undesirable products chromatographically by preparative HPLC.

The structural elucidation of the product [13C6]-D3G was performed by NMR and especially bythe fragmentation pattern after LC-MS/MS measurements. For two reasons, a complete assignmentof all NMR signals was not feasible: (a) 13C-labeled compounds show intense signals and complexcoupling behavior of 1H and 13C during NMR analysis in combination with (b) the small amount of[13C6]-D3G (170 µg). However, the 1H-NMR signals of H-10, H-11, and H-3, as well as the protons ofthe methyl groups at C-16 and C-14 of [13C6]-D3G could be clearly assigned. They showed similarshifts as Blackwell et al. [42] have already reported for DON. The 13C-NMR experiment revealed theexpected shifts of the keto-group C-8 at 202.2 ppm, the olefinic carbon C-10 resonating at 137.0 ppmas well as the shifts of the carbon atoms C-9, C-13, C-15, and C-14 of the DON structure. The glucosemoiety could be easily assigned by the characteristic coupling pattern of the signals with typical1Jcc coupling constants between 42 and 46 Hz. By means of 2D-experiments (HMBC optimized for2,3JH,C couplings) the key correlation signal between C-11 (97.4 ppm) and H-3 (4.54 ppm) could beobserved and, hence, the glycosidic bond was verified via NMR (Figure 4). For application as anisotopically-labeled standard, the unequivocal quantitation of [13C6]-D3G via qNMR of the resolvedsignal of H-10 provided the most accurate quantitation of such small amounts.

Molecules 2016, 21, 838 8 of 15

Molecules 2016, 21, 838 8 of 15

sensitivity would not have been achieved with our instrumentation when using a dilute-and-shoot approach as described by Al-Taher et al., Hu et al., and Malachova et al. [46–48].

To be able to expand the method to a multi-mycotoxin stable isotope dilution method including non-polar Fusarium toxins, like type A trichothecenes, zearalenone, and enniatins, a SPE-cartridge without activated carbon was used. Otherwise, non-polar mycotoxins would adsorb onto multifunctional SPE-cartridges usually containing activated carbon. Likewise to be able to include other mycotoxins in the developed method the samples were reconstituted with acetonitrile/water in a volume ratio of 1 to 1.

Figure 4. Section of the HMBC NMR spectrum of [13C6]-D3G showing indicating the glycosidic bond of 13C-1′ from [13C6]-glucose to H-3 of DON.

3.3. Method Validation

The LODs and LOQs of the presented method for D3G (2.99 μg/L and 8.84 μg/L) and DON (1.44 μg/L and 4.44 μg/L) were similar to those reported by Kostelanska et al. [43] in beer matrix and Nathanail et al. [20] in cereals. Thus, the sensitivity of our method is sufficiently low to quantitate D3G and DON in natural contaminated beers despite of the small sample volume of 0.8 mL. The intra-day and inter-day precisions, with maximum relative standard deviations of 0.5% and 7%, respectively, were similar or even better to previous studies [43,48]. As expected for SIDAs, the recoveries of the newly developed SIDA for D3G and DON ranged between 97% and 112% and, hence, were close to 100%. Due to the previous lack of a stable isotopologue, D3G was often quantified via matrix-matched calibration resulting in rather dissatisfactory recoveries between 39%

Figure 4. Section of the HMBC NMR spectrum of [13C6]-D3G showing indicating the glycosidic bondof 13C-11 from [13C6]-glucose to H-3 of DON.

3.2. Method Development

As already described in Habler and Rychlik [18], D3G had the same fragmentation pattern asDON due to in-source fragmentation and loss of glucose from D3G in the positive ESI mode. Therefore,a chromatographic baseline separation of these analytes was essential to allow quantification of DONinstead of the sum of D3G and DON in the positive ESI mode. This was achieved using a HydrosphereRP-C18 column (YMC Europe GmbH, Dinslaken, Germany) as the stationary phase combined witha moderately shallow gradient of 11.25% increase of acetonitrile per minute. To enable optimal MSconditions and sensitivity for both analytes, two chromatographic runs were performed for eachsample: The first one was run for D3G in the negative ESI mode without additives in the mobile phaseand the second one for DON in the positive ESI mode with formic acid in the mobile phase.

Sample purification using solid phase extraction was designed to reduce consumption of labeledinternal standard and to maximize analyte sensitivity. Moreover, a clean-up step is important toextend the HPLC column lifetime and to maintain performance of the ESI-MS source. After additionof acetonitrile to beer samples interfering polar matrix compounds can be precipitated as is alreadyoften applied [23,43–45]. However, a precipitation of polar mycotoxins, along with the matrix and aresulting poor recovery rate is avoided by using a small sample volume of 0.8 mL beer combined withusing the isotopologues [13C6]-D3G and [13C15]-DON as internal standards. A sufficient sensitivity

Molecules 2016, 21, 838 9 of 15

would not have been achieved with our instrumentation when using a dilute-and-shoot approach asdescribed by Al-Taher et al., Hu et al., and Malachova et al. [46–48].

To be able to expand the method to a multi-mycotoxin stable isotope dilution methodincluding non-polar Fusarium toxins, like type A trichothecenes, zearalenone, and enniatins, aSPE-cartridge without activated carbon was used. Otherwise, non-polar mycotoxins would adsorbonto multifunctional SPE-cartridges usually containing activated carbon. Likewise to be able to includeother mycotoxins in the developed method the samples were reconstituted with acetonitrile/water ina volume ratio of 1 to 1.

3.3. Method Validation

The LODs and LOQs of the presented method for D3G (2.99 µg/L and 8.84 µg/L) and DON(1.44 µg/L and 4.44 µg/L) were similar to those reported by Kostelanska et al. [43] in beer matrix andNathanail et al. [20] in cereals. Thus, the sensitivity of our method is sufficiently low to quantitate D3Gand DON in natural contaminated beers despite of the small sample volume of 0.8 mL. The intra-dayand inter-day precisions, with maximum relative standard deviations of 0.5% and 7%, respectively,were similar or even better to previous studies [43,48]. As expected for SIDAs, the recoveries of thenewly developed SIDA for D3G and DON ranged between 97% and 112% and, hence, were close to100%. Due to the previous lack of a stable isotopologue, D3G was often quantified via matrix-matchedcalibration resulting in rather dissatisfactory recoveries between 39% and 117% [18,43,48]. Thusthe validation data of the stable isotope dilution LC-MS/MS method for D3G and DON proved itssuitability for further applications.

3.4. Analysis of D3G and DON in Beer Samples

The newly-developed SIDA for D3G and DON was firstly applied to beer samples from differentcountries (Germany, Taiwan, China, USA) with varying alcohol contents between 4.7% and 8.3%. Onaverage, 78% and 89% of the investigated beer samples were found to be contaminated, in detail with22.6 µg/L D3G and 13.3 µg/L DON, respectively. Except for one IPA from the US, the contents ofD3G and DON were rather low and at a maximum of 20 µg/L, respectively. This IPA with the highestalcohol content of 8.3% revealed 63.3 µg/L D3G and 28.3 µg/L DON. The maximum permitted contentof D3G and DON in beer has to be derived from the legislatively-set limit for DON in unprocessedcereals of 1250 µg/kg, that is applicable to brewing malt [3]. The maximum permitted content ofDON in beer would be 225 µg/L (1250 µg/kg ˆ 18%) calculated by the charge of normal lager. Thecontents of the investigated beer samples were at least five times below that limit for D3G (calculatedas DON) and eight times for DON. Due to the limited number of analyzed beer samples, our studydoes not allow a risk assessment concerning organic and conventional beers or a relation betweenalcohol content and contamination level like Harcz et al. [49] and Kostelanska et al. [43] did. Otherstudies that investigated D3G and DON in beer revealed similar incidences and contents for D3Gbetween 1.2 µg/L and 89.3 µg/L and, for DON, between 1.0 µg/L and 81.3 µg/L [24,43,50]. In ourstudy, D3G and DON could often be detected together. The molar ratios of D3G/DON were between0.34 and 1.54, but mostly above 1.0, which implies that the levels of D3G exceed the levels of DON. Incontrast, Varga et al. [24] and Kostelanska et al. [43] reported molar ratios of D3G/DON between 0.11and 1.25 (averaged 0.56).

This newly-developed LC-MS/MS method for D3G and DON based on a SIDA provides an easyapproach to analyze all kinds of beer. With the synthesized labeled standard of D3G a manifold ofbeer samples, as well as barley malt and other cereals, can be investigated via SIDA in prospectivestudies. This is necessary to gain exact and sufficient exposure and toxicity data to be able to derive ajustified maximum limit for D3G in food and feed.

Molecules 2016, 21, 838 10 of 15

4. Materials and Methods

4.1. Chemicals and Reagents

Acetonitrile, water (both analytical grade), diethyl ether, hydrochloric acid, potassium thiocyanate,sodium chloride, sodium hydrogen carbonate, sodium sulfate, and sodium thiosulfate were purchasedfrom VWR (Ismaning, Germany). DON, formic acid (>95%), silver carbonate, hexane, pyridine, aceticanhydride, hydrogen bromide in acetic acid (33 wt %), deuterium oxide, and deuterated chloroformwere bought from Sigma Aldrich (Steinheim, Germany). Dichloromethane was purchased from Merck(Darmstadt, Germany) and [13C6]-glucose was aquired from Euriso-Top (Saarbrücken, Germany). Theunlabeled reference compounds D3G and DON, as well as the labeled standard [13C15]-DON, werebought from Coring System Diagnostix (Gernsheim, Germany).

4.2. Syntheses

4.2.1. Synthesis of 2,3,4,6-Tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside

2.65 mL acetic anhydride (2.85 g, 28 mmol, 10 eq.) and 2.25 mL pyridine (2.20 g, 28 mmol, 10 eq.)were added to [13C6]-glucose (0.52 g, 2.8 mmol, 1 eq.). The mixture was stirred at room temperaturefor 48 h until the solution was clear. The solution was poured in 10 mL ice cold water and the productwas extracted three times with 10 mL dichloromethane, respectively. The combined organic phaseswere washed with 10 mL brine and dried over sodium sulfate. Dichloromethane was evaporated and1,2,3,4,6-pentaacetyl-D-[13C6]-glucopyranoside was stored at ´20 ˝C in the dark.

While cooling, 2.5 mL hydrogen bromide (33 wt % in acetic acid, 5 mL/g glucose) were addedto 1,2,3,4,6-pentaacetyl-D-[13C6]-glucopyranoside (0.5 g, 1.3 mmol). Until the solution was clear,stirring was performed under cooling. Afterwards, the solution was stirred at room temperaturefor further 6 h. The solution was poured on 10 mL ice cold water and extracted three times with10 mL dichloromethane, respectively. The combined organic phases were washed once with 10 mLsodium thiosulfate solution (0.5%), three times with 10 mL saturated sodium hydrogen carbonatesolution, respectively, and once with 10 mL brine. The organic phase was dried over sodium sulfateand evaporated to dryness. The residue was solved in diethyl ether and overlaid with hexane. Aftercrystallizing under cooling, the solvents were removed under a gentle nitrogen stream and the productwas stored at ´20 ˝C in the dark. After conducting these syntheses four times, NMR measurements indeuterated chloroform were performed.

4.2.2. Synthesis of Deoxynivalenol-3-β-D-[13C6]-glucoside

DON (0.25 mg, 0.85 µmol) was dissolved in 0.3 mL dichloromethane. To the solution2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside (0.007 g, 17 µmol) and silver carbonate(0.004 g, 15 µmol) were added and stirred at room temperature in the dark. Another portions of2,3,4,6-tetraacetyl-1-bromo-α-D-[13C6]-glucopyranoside (0.007 g, 17 µmol) and silver carbonate (0.004 g,15 µmol) were added two times after 24 h, respectively. The reaction was stirred for a further three daysat room temperature in the dark. The solution was membrane filtrated (0.45 µm) and the solvent wasremoved under a nitrogen stream at 50 ˝C. Then the synthesis was reconstituted with 1 mL potassiumthiocyanate solution (0.5%) to precipitate surplus silver ions. After membrane filtration (0.45 µm)2.5 mL ACN and 1.5 mL H2O were added to the solution of the protected target compound.

For hydrolysis of DON-2,3,4,6-tetra-acetyl-[13C6]-glucoside to give DON-3-[13C6]-glucosidepotassium hydroxide (0.012 g, 215 µmol) was dissolved in 0.5 mL H2O and was added dropwiseunder pH control by pH meter to the DON-2,3,4,6-tetra-acetyl-[13C6]-glucoside solution. The pH wasnot allowed to exceed 11. The reaction mixture was stored overnight at 5 ˝C and was neutralizedthereafter with 1 M and 0.1 M hydrochloride acid (around four drops each). The synthesis procedurewas performed 72 times.

Molecules 2016, 21, 838 11 of 15

The product DON-3-[13C6]-glucoside was separated from DON-15-[13C6]-glucoside, DON, andother side products and purified via preparative HPLC. After combining and removing the solvent ofthese fractions containing DON-3-[13C6]-glucoside NMR and qNMR experiments in deuterium oxide,as well as LC-MS/MS measurements were performed.

4.3. Preparative HPLC

Preparation and purification of DON-3-[13C6]-glucoside was carried out on a Merck Hitachisystem (Kyoto, Japan) consisting of two pumps (L-7100, Merck Hitachi, Tokyo, Japan), a degasser(Degassex model DG 4400, Phenomenex, Aschaffenburg, Germany), an autosampler (L-7200, MerckHitachi, Tokyo, Japan), and a UV detector (L-7400, Merck Hitachi, Tokyo, Japan). As the stationaryphase a Pro Pack RP-C18 column (150 mm ˆ 10 mm, S-5 µm, 12 nm, YMC Europe GmbH, Dinslaken,Germany) combined with a widepore C18-guard column (4 mmˆ 3 mm, Security Guard, Phenomenex,Aschaffenburg, Germany) was used. The binary gradient system consisted of (A) water and (B)acetonitrile at a flow rate of 1 mL/min. The shallow gradient to separate DON-3-[13C6]-glucoside fromDON-15-[13C6]-glucoside, DON, and other side products was raised linearly from 10% B to 100% Bduring the next 20 min, and then maintained at 100% B for 5 min. Next, the mobile phase returned to10% B within 2 min and the system was equilibrated for 10 min before the next run. The detection wasperformed at 218 nm and the injection volume was 100 µL.

4.4. NMR and qNMR

The NMR and qNMR measurements were performed on a Bruker AV III system (400 MHz,Bruker, Rheinstetten, Germany) and on a Bruker AV III system (500 MHz, Bruker, Rheinstetten,Germany), respectively. 1H- and 13C-NMR spectra were recorded at 400 or 500 MHz and at 100.61 or125.76 MHz, respectively. Data acquisition was performed with TopSpin software 3.2 (Bruker,Rheinstetten, Germany) and MestReNova software 10.0 (Mestrelab Research, La Coruña, Spain).The samples were dissolved in 600 µL or 300 µL deuterium oxide or deuterated chloroform. TheqNMR measurement of DON-3-[13C6]-glucoside was performed as published by Korn et al. [34]. Inbrief, the compound was dissolved in 600 µL deuterium oxide and analyzed in 5ˆ 178 mm NMR tubes(Bruker, Rheinstetten, Germany). As an external standard a caffeine sample with known concentrationwas used.

4.5. Preparation of Standard Solution

Stock solutions of labeled and unlabeled toxins were prepared in a concentration of 10 µg/mL inacetonitrile and further diluted to a final concentration of 1 µg/mL. All solutions were stored at 4 ˝Cin the dark.

4.6. LC-MS/MS

LC-MS/MS was carried out on a Shimadzu LC-30A Prominence system (Shimadzu, Kyoto, Japan)with a Hydrosphere RP-C18 column (150 mmˆ 3.0 mm, S-3 µm, 12 nm, YMC Europe GmbH, Dinslaken,Germany) and a C18-guard column (Phenomenex, Aschaffenburg, Germany) as the stationary phasethat was kept at 40 ˝C. The binary gradient system consisted of (A) water and (B) acetonitrile forD3G and of (A) 0.1% formic acid and (B) acetonitrile with 0.1% formic acid for DON at a flow rate of0.2 mL/min. The gradient for D3G was started and held at 10% B for 1 min, raised linearly from 10% Bto 100% B during the next 6 min, and then maintained at 100% B for 4 min. Next, the mobile phasereturned to 10% B within 2 min and the system was equilibrated for 10 min before the next run. Thegradient for DON was started and held at 10% B for 1 min, raised linearly from 10% B to 100% Bduring the next 8 min, and then maintained at 100% B for 4 min. Next, the mobile phase returnedto 10% B within 2 min and the system was equilibrated for 10 min before the next run. The injectionvolume was 5 µL.

Molecules 2016, 21, 838 12 of 15

The LC was interfaced with a triple quadrupole ion trap mass spectrometer (LCMS-8050,Shimadzu, Kyoto, Japan). It operated in the negative ESI mode for the analyte D3G and the internalstandard [13C6]-D3G and in the positive ESI mode for the analyte DON and the internal standard[13C15]-DON. The ion source parameters were set as follows: interface temperature 300 ˝C, heatblock temperature 400 ˝C, dilution line temperature 250 ˝C, heating gas flow 10 L/min, drying gasflow 10 L/min, nebulizing gas flow 3 L/min, collision-induced dissociation gas 17 kPa, interfacevoltage 4 kV. MS parameters were optimized by injection of each standard solution (1 µg/mL).The mass spectrometer was operated in the multiple reaction monitoring (MRM) mode for MS/MSmeasurements at the conditions detailed in Table 1. A waste valve diverted the column effluent to themass spectrometer from 6 to 10 min in both modes. Data acquisition was performed with LabSolutionssoftware 5.80 (Shimadzu, Kyoto, Japan).

4.7. Calibration and Quantitation

For the response curves constant amounts of internal standard (S) were mixed with varyingamounts of analyte (A) in molar ratios between 0.1 and 10 (1:10, 1:5, 1:2, 1:1; 2:1, 5:1, 10:1). Theconcentrations ranged for D3G between 0.003 µg/L and 0.3 µg/L and for DON between 0.002 µg/Land 0.2 µg/L. All analytes applying SIDA had [13C]-labeled isotopologues. The response curveswere calculated from molar ratios [n(S)/n(A)] versus peak area ratios [A(S)/A(A)] after LC-MS/MSmeasurement. Response functions were obtained using linear regression.

4.8. Sample Preparation

To 0.8 mL of degassed beer 4 mL acetonitrile and the internal standards [13C6]-D3G (30 µL of1 µg/mL) and [13C15]-DON (20 µL of 1 µg/mL) were added. The samples were vortexed (20 s) andwere completely applied on a Bond Elut Mycotoxin cartridge (500 mg, 3 mL, Agilent Technologies,Santa Clara, CA, USA). The liquids were passed through the cartridges by vacuum suction. Theeluates were collected and evaporated until dryness. The samples were reconstituted with 200 µLacetonitrile/water (1/1, v/v) and membrane filtered (0.45 µm). Afterwards, LC-MS/MS analysiswas performed.

4.9. Method Validation

4.9.1. Limits of Detection and Quantitation

Beer free of the monitored mycotoxins was chosen as the blank matrix and used to determineLODs and LOQs as suggested by Vogelgesang and Hädrich [38]. Therefore, the blank matrix wasspiked with the unlabeled analytes at four different amounts (3, 10, 20, and 30 µg/L for D3G andDON), each in triplicate.

4.9.2. Recovery

Blank beer was spiked in triplicate with different amounts of analytes (10, 20, and 30 µg/L forD3G and DON) and analyzed as described previously. Recoveries were calculated as the ratio ofdetected and spiked contents.

4.9.3. Precision

Naturally-contaminated beer with contents of D3G at 28 µg/L and DON at 15 µg/L was used forintra-day (n = 3) and inter-day (n = 3, within two weeks) precision measurements.

Acknowledgments: This project was supported by the Forschungskreis der Ernährungsindustrie e.V. (FEIBonn), the AiF, the German Federal Ministry of Economic Affairs and Energy (AiF-Project No.: 17221 N),Wissenschaftsförderung der Deutschen Brauwirtschaft e.V., and Wissenschaftliche Station für Brauerei in Münchene.V.. Moreover, the authors gratefully acknowledge the support by the Chair of Food Chemistry and MolecularScience, Technische Universität München, Germany, for the NMR and qNMR measurements.

Molecules 2016, 21, 838 13 of 15

Author Contributions: K.H. and M.R. conceived and designed the experiments; K.H. performed the experiments;O.F. conducted the NMR experiments; K.H. and O.F. analyzed the data; K.H. wrote the paper.

Conflicts of Interest: The authors declare no conflict of interest. The founding sponsors had no role in the designof the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in thedecision to publish the results.

Abbreviations

The following abbreviations are used in this manuscript:

D3G Deoxynivalenol-3-glucosideDON DeoxynivalenolESI Electrospray IonizationHMBC Heteronuclear Multiple Bond CorrelationLC-MS/MS Liquid chromatography combined with tandem mass spectrometryLOD Limit of DetectionLOQ Limit of QuantitationNMR Nuclear Magnetic ResonanceqNMR quantitative NMRSIDA stable isotope dilution assaySPE solid phase extraction

References

1. Parry, D.W.; Jenkinson, P.; McLeed, L. Fusarium ear blight (scab) in small grain cereals—A review.Plant Pathol. 1995, 44, 207–218. [CrossRef]

2. Desjardins, A.E. Fusarium Mycotoxins: Chemistry, Genetics and Biology; APS Press: Saint Paul, MN, USA, 2006.3. The Commission of the European Communities. Commission Regulation (EC) No. 1881/2006 of 19 December

2006 setting maximum levels for certain contaminants in foodstuffs; Official Journal of the European Union,European Union: Brussels, Belgium, 2006.

4. Food and Agricultural Organization of the United Nations/World Health Organization (FAO/WHO). Safetyevaluation of certain contaminants in food/prepared by the Seventy-second meeting of the Joint FAO/WHOExpert Committee on Food Additives (JECFA). In WHO Food Additives Series No. 63/FAO JECFA Monographs8; World Health Organization: Geneva, Switzerland and Rome, Italy, 2011.

5. Sewald, N.; von Gleissenthall, J.L.; Schuster, M.; Müller, G.; Aplin, R.T. Structural Elucidation of a PlantMetabolite of 4-Deosoxynivalenol. Tetrahedron 1992, 3, 953–960. [CrossRef]

6. Berthiller, F.; Dall’Asta, C.; Schuhmacher, R.; Lemmens, M.; Adam, G.; Krska, R. Masked Mycotoxins:Determination of a Deoxynivalenol Glucoside in Artificially and Naturally Contaminated Wheat by LiquidChromatography-Tandem Mass Spectrometry. J. Agric. Food Chem. 2005, 53, 3421–3425. [CrossRef] [PubMed]

7. Nagl, V.; Schwartz, H.; Krska, R.; Moll, W.-D.; Knasmüller, S.; Ritzmann, M.; Adam, G.; Berthiller, F.Metabolism of the masked mycotoxin deoxynivalenol-3-glucoside in rats. Toxicol. Lett. 2012, 213, 367–373.

8. Nagl, V.; Woechtl, B.; Schwartz-Zimmermann, H.E.; Hennig-Pauka, I.; Moll, W.-D.; Adam, G.; Berthiller, F.Metabolism of the masked mycotoxin deoxynivalenol-3-glucoside in pigs. Toxicol. Lett. 2014, 229, 190–197.

9. Berthiller, F.; Krska, R.; Domig, K.J.; Kneifel, W.; Juge, N.; Schuhmacher, R.; Adam, G. Hydrolytic fate ofdeoxynivalenol-3-glucoside during digestion. Toxicol. Lett. 2011, 206, 264–267. [CrossRef] [PubMed]

10. Rychlik, M.; Humpf, H.-U.; Marko, D.; Dänicke, S.; Mally, A.; Berthiller, F.; Klaffke, H.; Lorenz, N. Proposalof a comprehensive definition of modified and other forms of mycotoxins including “masked” mycotoxins.Mycotoxin Res. 2014, 30, 197–205. [CrossRef] [PubMed]

11. Berthiller, F.; Crews, C.; Dall’Asta, C.; Saeger, S.D.; Haesaert, G.; Karlovsky, P.; Oswald, I.P.; Seefelder, W.;Speijers, G.; Stroka, J. Masked mycotoxins: A review. Mol. Nutr. Food Res. 2013, 57, 165–186. [CrossRef]

12. Zachariasova, M.; Vaclavikova, M.; Lacina, O.; Vaclavik, L.; Hajslova, J. Deoxynivalenol Oligoglycosides:New “Masked” Fusarium Toxins Occurring in Malt, Beer, and Breadstuff. J. Agric. Food Chem. 2012, 60,9280–9291. [CrossRef] [PubMed]

13. Gardiner, S.A.; Boddu, J.; Berthiller, F.; Hametner, C.; Stupar, R.M.; Adam, G.; Muehlbauer, G.J. TranscriptomeAnalysis of the Barley-Deoxynivalenol Interaction: Evidence for a Role of Glutathione in DeoxynivalenolDetoxification. Am. Phytopathol. Soc. 2010, 23, 962–976. [CrossRef] [PubMed]

Molecules 2016, 21, 838 14 of 15

14. Kluger, B.; Bueschl, C.; Lemmens, M.; Michlmayr, H.; Malachova, A.; Koutnik, A.; Maloku, I.; Berthiller, F.;Adam, G.; Krska, R.; et al. Biotransformation of the Mycotoxin Deoxynivalenol in Fusarium Resistant andSusceptible Near Isogenic Wheat Lines. PLoS ONE 2015, 10, e0119656. [CrossRef] [PubMed]

15. Warth, B.; Fruhmann, P.; Wiesenberger, G.; Kluger, B.; Sarkanj, B.; Lemmens, M.; Hametner, C.; Fröhlich, J.;Adam, G.; Krska, R.; et al. Deoxynivalenol-sulfates: Identification and quantification of novel conjugated(masked) mycotoxins in wheat. Anal. Bioanal. Chem. 2015, 407, 1033–1039. [CrossRef] [PubMed]

16. Yoshinari, T.; Sakuda, S.; Furihata, K.; Furusawa, H.; Ohnishi, T.; Sugita-Konishi, Y.; Ishizaki, N.; Terajima, J.Structural determination of a nivalenol glucoside and development of an analytical method for thesimultaneous determination of nivalenol and deoxynivalenol, and their glucosides, in wheat. J. Agric.Food Chem. 2014, 62, 1174–1180. [CrossRef] [PubMed]

17. Stanic, A.; Uhlig, S.; Solhaug, A.; Rise, F.; Wilkins, A.L.; Miles, C.O. Preparation and Characterization ofCysteine Adducts of Deoxynivalenol. J. Agric. Food Chem. 2016. in press.

18. Habler, K.; Rychlik, M. Multi-mycotoxin stable isotope dilution LC-MS/MS method for Fusarium toxins incereals. Anal. Bioanal. Chem. 2016, 408, 307–317. [CrossRef] [PubMed]

19. Habler, K.; Hofer, K.; Geißinger, C.; Schüler, J.; Hückelhoven, R.; Hess, M.; Gastl, M.; Rychlik, M. Fate ofFusarium Toxins during the Malting Process. J. Agric. Food Chem. 2016, 64, 1377–1384. [CrossRef] [PubMed]

20. Nathanail, A.V.; Syvähuoko, J.; Malachová, A.; Jestoi, M.; Varga, E.; Michlmayr, H.; Adam, G.; Sieviläinen, E.;Berthiller, F.; Peltonen, K. Simultaneous determination of major type A and B trichothecenes, zearalenoneand certain modified metabolites in Finnish cereal grains with a novel liquid chromatography-tandem massspectrometric method. Anal. Bioanal. Chem. 2015, 407, 4745–4755. [CrossRef] [PubMed]

21. Kostelanska, M.; Zachariasova, M.; Dzuman, Z.; Hajslova, J.; Ehrenbergerova, J.; Cerkal, R.; Vaculova, K.;Mikyska, A.; Psota, V. Fusarium Mycotoxins in Spring Barley and Their Occurrence Within the TechnologicalChain Barley-Malt-Beer. Kvas. Prumysl 2011, 57, 209–214.

22. Malachova, A.; Dzuman, Z.; Veprikova, Z.; Vaclavikova, M.; Zachariasova, M.; Hajslova, J. Deoxynivalenol,Deoxynivalenol-3-glucoside, and Enniatins: The Major Mycotoxins Found in Cereal-Based Products on theCzech Market. J. Agric. Food Chem. 2011, 59, 12990–12997. [CrossRef] [PubMed]

23. Lancova, K.; Hajslova, J.; Poustka, J.; Krplova, A.; Zachariasova, M.; Dostalek, P.; Sachambula, L. Transfer ofFusarium mycotoxins and ‘masked’ deoxynivalenol (deoxynivalenol-3-glucoside) from field barley throughmalt to beer. Food Addit. Contam. 2008, 25, 732–744. [CrossRef] [PubMed]

24. Varga, E.; Malachova, A.; Schwartz, H.; Krska, R.; Berthiller, F. Survey of deoxynivalenol and its conjugatesdeoxynivalenol-3-glucoside and 3-acetyl-deoxynivalenol in 374 beer samples. Food Addit. Contam. Part A2013, 30, 137–146. [CrossRef] [PubMed]

25. Klötzel, M.; Gutsche, B.; Lauber, U.; Humpf, H.-U. Determination of 12 type A and B trichothecenes incereals by liquid chromatography-electrospray ionization tandem mass spectrometry. J. Agric. Food Chem.2005, 53, 8904–8910. [CrossRef] [PubMed]

26. Asam, S.; Rychlik, M. Synthesis of Four Carbon-13-Labeled Type A Trichothecene Mycotoxins and TheirApplication as Internal Standards in Stable Isotope Dilution Assays. J. Agric. Food Chem. 2006, 54, 6535–6546.

27. Ravindranathan Kartha, K.P.; Jennings, H.J. A Simplified, One-Pot Preparation of Acetobromosugars fromReducing Sugars. J. Carbohydr. Chem. 1990, 9, 777–781. [CrossRef]

28. Redemann, E.C.; Niemann, C. Acetobromoglucose: [2,3,4,6-Tetraacetyl-α-D-glucopyranosyl bromide].Org. Synth. 1955, 3, 11.

29. Koschella, A.; Dorn, S.; Heinze, T.; Togola, A.; Paulsen, B.S. Synthesis and characterization of branchedpolysaccharides by reaction of cellulose with 2,3,4,6-tetraacetyl-1-bromo-α-Dglucopyranoside. Arkivoc 2012,3, 76–89.

30. Savard, M.E. Deoxynivalenol Fatty Acid and Glucoside Conjugates. J. Agric. Food Chem. 1991, 39, 570–574.31. Mikula, H.; Hametner, C.; Berthiller, F.; Warth, B.; Krska, R.; Adam, G.; Fröhlich, J. Fast and reproducible

chemical synthesis of zeralenone-14-beta, D-glucuronide. World Mykotoxin J. 2012, 5, 289–296. [CrossRef]32. Mikula, H.; Weber, J.; Lexmüller, S.; Bichl, G.; Schwartz, H.; Varga, E.; Berthiller, F.; Hametner, C.; Krska, R.;

Fröhlich, J. Simultaneous preparation of α/β-zearalenol glucosides and glucuronides. Carbohydr. Res. 2013,373, 59–63. [CrossRef] [PubMed]

33. Malachová, A.; Štocková, L.; Wakker, A.; Varga, E.; Krska, R.; Michlmayr, H.; Adam, G.; Berthiller, F. Criticalevaluation of indirect methods for the determination of deoxynivalenol and its conjugated forms in cereals.Anal. Bioanal. Chem. 2015, 407, 6009–6020. [CrossRef] [PubMed]

Molecules 2016, 21, 838 15 of 15

34. Korn, M.; Frank, O.; Hofmann, T.; Rychlik, M. Development of stable isotope dilution assays for ochratoxinA in blood samples. Anal. Biochem. 2011, 419, 88–94. [CrossRef] [PubMed]

35. Berthiller, F.; Schuhmacher, R.; Buttinger, G.; Freudenschuss, M.; Adam, G.; Krska, R. Synthesis ofDeoxynivalenol-Glucosides and their Characerization using a QTrap LC-MS/MS. Mycotoxin Res. 2003,19, 47–50. [CrossRef] [PubMed]

36. Agilent Technologies. New SPE Sorbent for Clean-up of Fusarium Toxin-Contaminated Cereals &Cereal-based Foods, Bond Elut Mycotoxin. 2010. Available online: https://www.agilent.com/cs/library/applications/SI-00295.pdf (accessed on 3 May 2016).

37. Agilent Technologies. LC/MS/MS of Trichothecenes and Zearalenone in Wheat Using Different SamplePrep Methods. 2012. Available online: https://www.agilent.com/cs/library/applications/5990-9107EN.pdf(accessed on 3 May 2016).

38. Vogelgesang, J.; Hädrich, J. Limits of detection, identification and determination: A statistical approach forpractitioners. Accredit. Qual. Assur. 1998, 3, 242–255. [CrossRef]

39. Koenigs, W.; Knorr, E. Ueber einige Derivate des Traubenzuckers und der Galactose. Ber. Dtsch. Chem. Ges.1901, 34, 957–981. [CrossRef]

40. Schmidt, R.R.; Michel, J. Facile Synthesis of α- and β-O-Glycosyl Imidates; Preparation of Glycosides andDisaccharides. Angew. Chem. Int. Ed. Engl. 1980, 19, 731–732. [CrossRef]

41. Van Beyersbergen Henegouwen, W.G.; Hiemstra, H. Studies toward the Total Synthesis of the OxindoleAlkaloid Gelsedine: An Efficient Allene-Terminated N-Acyliminium Ion Cyclization. J. Org. Chem. 1997, 62,8862–8867. [CrossRef]

42. Blackwell, B.A.; Greenhalgh, R.; Bain, A.D. Carbon-13 and proton nuclear magnetic resonance spectralassignments of deoxynivalenol and other mycotoxins from Fusarium graminearum. J. Agric. Food Chem.1984, 32, 1078–1083. [CrossRef]

43. Kostelanska, M.; Hajslova, J.; Zachariasova, M.; Malachova, A.; Kalachova, K.; Poustka, J.; Fiala, J.; Scott, P.M.;Berthiller, F.; Krska, R. Occurrence of Deoxynivalenol and Its Major Conjugate, Deoxynivalenol-3-Glucoside,in Beer and Some Brewing Intermediates. J. Agric. Food Chem. 2009, 57, 3187–3194. [CrossRef] [PubMed]

44. Zachariasova, M.; Cajka, T.; Godula, M.; Malachova, A.; Veprikova, Z.; Hajslova, J. Analysis of multiplemycotoxins in beer employing (ultra)-high-resolution mass spectrometry. Rapid Commun. Mass Spectrom.2010, 24, 3357–3367. [CrossRef] [PubMed]

45. Vaclavikova, M.; Malachova, A.; Veprikova, Z.; Dzuman, Z.; Zachariasova, M.; Hajslova, J. ‘Emerging’mycotoxins in cereals processing chains: Changes of enniatins during beer and bread making. Food Chem.2013, 136, 750–757. [CrossRef] [PubMed]

46. Al-Taher, F.; Banaszewski, K.; Jackson, L.; Zweigenbaum, J.; Ryu, D.; Cappozzo, J. Rapid Method for theDetermination of Multiple Mycotoxins in Wines and Beers by LC-MS/MS Using a Stable Isotope DilutionAssay. J. Agric. Food Chem. 2013, 61, 2378–2384. [CrossRef] [PubMed]

47. Hu, L.; Gastl, M.; Linkmeyer, A.; Hess, M.; Rychlik, M. Fate of enniatins and beauvericin during the maltingand brewing process determined by stable isotope dilution assays. LWT Food Sci. Technol. 2014, 56, 469–477.

48. Malachova, A.; Varga, E.; Schwartz, H.; Krska, R.; Berthiller, F. Development, validation and application ofan LC-MS/MS based method for the determination of deoxynivalenol and its conjugates in different typesof beer. World Mycotoxin J. 2012, 5, 261–270. [CrossRef]

49. Harcz, P.; Tangni, E.K.; Wilmart, O.; Moons, E.; van Peteghem, C.; de Saeger, S.; Schneider, Y.-J.; Larondelle, Y.;Pussemier, L. Intake of ochratoxin A and deoxynivalenol through beer consumption in Belgium. Food Addit.Contam. 2007, 24, 910–916. [CrossRef] [PubMed]

50. Zachariasova, M.; Hajslova, J.; Kostelanska, M.; Poustka, J.; Krplova, A.; Cuhra, P.; Hochel, I. Deoxynivalenoland its conjugates in beer: A critical assessment of data obtained by enzyme-linked immunosorbent assayand liquid chromatography coupled to tandem mass spectrometry. Anal. Chim. Acta 2008, 625, 77–86.

Sample Availability: Samples of the compounds are not available from the authors.

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open accessarticle distributed under the terms and conditions of the Creative Commons Attribution(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).